Abstract
Executive Summary
Early marriage presents significant risks to the well-being of children around the world. Likewise, migrant children experience significant vulnerabilities in the United States as they confront complex legal and administrative proceedings. At a crucial intersection of migration and marriage, scholars, advocates, and lawmakers have recognized ways that US immigration law enables child marriage.
Research and public policy has yet to consider, however, a unique population at the juncture of marriage, migration, and childhood: wedded girls in immigration detention. Using administrative data that the Office of Refugee Resettlement provided in response to Freedom of Information Act requests and litigation, this study documents for the first time the reality of minor wives — married girls — in ORR immigration detention and the US government practice of releasing them to adult spouses and informal partners. The decision to release any minor from government custody to an adult sexual partner merits critical attention.
This study provides that initial critical attention and sets the stage for further research. Using a UNICEF definition of child marriage, this study considers both formal and informal unions. During the study period from October 1, 2014, to September 30, 2024, ORR released at least 427 minors to an adult sexual partner, 206 to a legal spouse and 221 to an informal partner. Because of data constraints that limit the identification of children in informal marriages, this study’s count likely underestimates the total number of children the US government has released from immigration detention to an adult sexual partner. Further, this study identifies and contextualizes risks for married migrant minors in the US states to which ORR has released the minors by reference to the states’ laws regarding the minimum age for marriage and statutory rape.
This article makes the following policy recommendations to reform the US detention system for migrant children: As ORR makes decisions regarding a minor’s release from immigration custody to an adult sexual partner, it should document its consideration of the risks and benefits in a mandatory, rather than discretionary, home study. If ORR decides to release a minor to an adult sexual partner, it should ensure that the release complies with relevant state law regarding marriage, sex, and consent. ORR should develop and “implement robust feedback mechanisms to ensure solicitation, documentation and consideration” of minors’ own views (Heidbrink and Díaz 2024, 20) to demonstrate respect for minors’ autonomy regarding their relationships despite the compounded constraints of childhood, immigration detention, and undocumented legal status. With access to additional qualitative data, further research could provide a greater understanding of ORR’s decision making processes than the quantitative data of this study allow. Further research should assess short, medium, and long-term outcomes for girls released from ORR immigration detention to their adult sexual partners, both to support girls’ well-being and to better inform ORR decision making.
. . . minor reported that she started a relationship with her current partner . . . at the age of 15 years old and he was 24. . . . she traveled to US with her partner and her son. . . . they were separated at the US border . . . . her partner is residing in CA with her son. . . . (SIR 235716) [17-year-old minor] states that this pregnancy is the result of a relationship with her 22-year-old common law husband in Guatemala. . . . (SIR 484451) [16-year-old minor] self-reports she is approximately 4 months pregnant. . . . Conception was consensual . . . with her husband, she says she is married. . . . (SIR 265455) [17-year-old] Minor’s pregnancy test was positive. . . . she has been married since April 2017, and had her first child on September, 2017. . . . (SIR 237996) [16-year-old minor] has been living with her common-law husband . . . for the past 18 months. [He] is currently 18 years old . . . . [He] has plans . . . in the near future to reunite[] with [her] and baby to continue to be a close family. (SIR 484428)
Early marriage presents significant risks to the well-being of children, especially girls, around the world (UNICEF 2023). 1 Child marriage complicates girls’ educational opportunities and attainment (McCleary-Sills et al. 2015; Psaki 2016), undermines their long-term economic well-being (Parsons et al. 2015; Wodon et al. 2017), and inhibits their physical and mental health (Nour 2009; Raj 2009; Fan and Koski 2022). Child marriage can also increase girls’ risk of intimate partner violence (Kidman 2017). Migrant children experience significant vulnerabilities in the United States as they confront complex legal proceedings, most often in a language other than their own and at times without an attorney or other legal representative (Luiselli 2017; Caswell and Ruehs-Navarro 2022; Galli and Padilla 2025). At a crucial intersection of migration and marriage, scholars, advocates, and lawmakers have recognized ways that US immigration law enables child marriage (Makhlouf 2017; United States Senate [US Senate] 2019; Hill and Elder 2020; Adu 2024; Bearat 2024).
The scholarship has yet to consider, however, a unique population at the juncture of marriage, migration, and childhood: wedded girls in immigration detention. In the United States, federal law charges the Office of Refugee Resettlement (ORR) with the custody and care of unaccompanied migrant children (6 USC § 279(g)(2), 8 USC § 1232(g)). Using administrative data ORR provided in response to Freedom of Information Act (FOIA) requests and litigation, this study documents for the first time the reality of minor wives — married girls — in ORR custody and considers the implications of marriage to their detention and release. When assessing to whom to release a married girl, should the United States government consider her adult husband or informal partner and thus recognize a child marriage? The decision to release any minor from government custody to an adult sexual partner merits critical attention, critical attention that is lacking in the scholarship and in US government policy.
With its focus on the release of unaccompanied girls to adult sexual partners, this research likewise contributes a missing piece to the growing scholarship on ORR’s decision making processes. Hannah Dreier’s 2023 reporting for the New York Times detailed the complicated realities of child labor exploitation and risk for some whom ORR released to unrelated individual sponsors (Dreier 2023a, 2023b). Marriage was outside the reporting’s scope. A 2024 comprehensive study completed by Lauren Heidbrink and Sarah J. Díaz (2024) provides detailed attention to the experiences of pregnant and parenting minors in ORR custody but does not consider marriage per se or that ORR may release children to adult sexual partners (pp. 81–85). The 2024 volume Kids in Cages edited by Emily Ruehs-Navarro, Lina Caswell Muñoz, and Sarah J. Díaz approaches ORR practices and children’s experiences in immigration custody from a rich variety of perspectives and methods but not marriage.
UNICEF (2023) defines child marriage as “any formal marriage or informal union between a child under the age of 18 years and an adult or another child.” Child marriage is a global phenomenon “with approximately one in five girls married in childhood” (UNICEF 2023). In 2015, the United Nations (2015) set a goal to eliminate early marriage worldwide by 2030. Since then, countries around the world have increasingly prohibited legal marriage involving children, although significant exceptions remain. Countries may also seek to regulate informal unions through statutory rape laws that criminalize sex with minors.
In the United States, individual states regulate formal marriage, not the federal government. Since the UN’s call to eliminate child marriage, US states have moved toward prohibitions on formal marriages involving children, but there is no consensus. Best estimates calculate that nearly 300,000 children entered formal marriages in the United States between 2000 and 2018 (Reiss 2021). Individual US states likewise regulate informal unions through statutory rape laws that criminalize sex with minors. The negative educational, economic, social, and health risks that child marriage creates for girls worldwide hold for the United States (Dahl 2010; Syrett 2016; Koski and Heyman 2018; Steinhus and Thompson 2020).
Unaccompanied Migrant Children: Definition and Legal Context
The United States Congress defines an “unaccompanied alien child” as a person who is (1) under 18 years of age, (2) without lawful immigration status in the United States, and (3) for whom no parent or legal guardian is in the United States or is “available to provide care and physical custody” (6 USC § 279(g)(2); 8 USC § 1232(g)).
Typically, officers from Customs and Border Protection (CBP) in the Department of Homeland Security (DHS) are the first government officials to encounter individuals who may be without lawful immigration status in the United States. CBP officers make an initial decision as to whether the individual is a minor and unaccompanied, although they do not distinguish between children they have separated from family members and those who arrive truly unaccompanied by an adult. When a married girl without lawful immigration status arrives to the United States with her husband but without a parent, she meets the legal definition of an unaccompanied migrant child. While marriage “has the power, legally and symbolically, to make people into adults” (Syrett 2016, 271), it does not do so for girls in immigration detention. Marriage does not make a migrant girl an adult ineligible for ORR custody.
Marriage is not, however, neutral with respect to an unaccompanied girl’s access to regular immigration status in the United States. If she is married to someone with legal status in the United States, and on a fact-specific basis, she may be able to regularize her status based on theirs (United States Citizenship and Immigration Services [USCIS] 2024). At the same time, a current marriage excludes her from Special Immigrant Juvenile consideration which is reserved for unmarried people under 21 who have been abused, abandoned, or neglected by a parent (USCIS 2026).
Federal law requires CBP to transfer unaccompanied children including married minors to ORR custody within 72 hours of apprehension (8 USC § 1232(b); United States Office of Refugee Resettlement [US ORR] 2025). ORR accepts unaccompanied children referred to it, and, in turn, places them with contracted private entities in locations across the United States. Children remain in ORR custody and in varying program placements (Alcaraz et al. 2024) until they no longer meet all components of the legal definition of an unaccompanied child.
The Trafficking Victims Protection Reauthorization Act of 2008 creates a mandatory legal structure and timeline for detained migrant children’s custody and release. Once ORR has received custody of a child, it must “promptly place[]” the child “in the least restrictive setting that is in the best interest of the child,” most often a shelter (8 USC § 1232(c)(2); Alcaraz et al. 2024). Additionally, for most of this study, the 1997 Flores Settlement Agreement set a detailed national standard for the care of detained migrant children. In April 2024, ORR promulgated as a formal regulation the Unaccompanied Children Foundational Rule which implemented portions of Flores including the requirement that ORR “release, without unnecessary delay” children in its custody “to a vetted and approved sponsor” (US ORR 2024, 34399).
The Flores Settlement did not explicitly contemplate a spouse as a sponsor to whom the government might release an unaccompanied minor from immigration detention. Both the Foundational Rule and ORR’s Unaccompanied Children Bureau Policy Guide, however, do. The Foundational Rule defines an unaccompanied child’s “close relative” to include an “immediate relative through legal marriage” (45 CFR § 410.1001). When identifying qualified sponsors to whom to release an unaccompanied child, ORR’s Policy Guide prioritizes (1) a parent or legal guardian, (2) an immediate family member, and (3) a more distant relative or unrelated person (US ORR 2025, § 2.2.1). Consistent with the Foundational Rule, the Policy Guide clarifies that Category 2 immediate family members include “relatives through legal marriage” (US ORR 2025, § 2.2.1). As discussed more fully below, during the study period, ORR categorized the general relationship between minors released to a formal spouse as Category 2 and the specific relationship as “UC’s Spouse,” meaning unaccompanied child’s spouse.
Data and Methods
Following multiple FOIA requests and enforcement litigation, ORR produced Excel spreadsheets of over 805,500 rows containing detailed demographic data on the 735,149 children who entered its custody in the 10 fiscal years between October 1, 2014, and September 30, 2024. 2 For each child, the primary dataset provides fields for: gender; age at ORR placement; country of birth; the US government agency referring the child to ORR custody; the referral date and time; initial ORR placement date and time; city and state of apprehension by the federal government; city and state of the child’s entry into the United States; ORR program placements by name, type, and admission date; and discharge categories and dates including internal transfers. For children released to an individual sponsor, a final discharge category termed Reunified (Individual Sponsor), the primary dataset includes the numerical representation of the sponsor’s general relationship with the child set forth in the Policy Guide’s priorities as well as a second more specific descriptor — mother, father, brother, sister, aunt, uncle, cousin, spouse, distant relative, unrelated sponsor, etc. The primary dataset does not include marital status. FOIA officers blinded the data by removing each child’s Alien Registration Number and ORR-assigned identification number but included a unique identifier local to its production of the primary dataset. The primary dataset allows longitudinal observations of a given anonymized child from the date of acceptance into ORR custody through final discharge.
In addition, this study draws on a secondary dataset. In December 2023, the New York Times made public and licensed for use, data regarding unaccompanied children in ORR custody that it had received from ORR also through FOIA requests and litigation (Dreier 2023c). The New York Times dataset includes children who entered ORR custody on or after January 1, 2015, and were discharged on or before May 26, 2023, to an individual sponsor (Dreier 2023c). The New York Times dataset overlaps with the primary dataset in fields for gender, country of origin, date of entry into ORR custody, and date of discharge for children released to an individual sponsor. Most importantly, the New York Times dataset includes the zip code of the individual sponsor to whom the child was released (Dreier 2023c). In this study’s FOIA litigation, ORR refused to provide all zip codes of individual sponsors to whom it had released children, despite having already done so to the New York Times for a shorter time period.
The illustrative epigraphs in this study are drawn from anonymized special incident reports (SIRs) that ORR produced for three 90-day periods: the first 90 days of the first Trump Administration (January 20, 2017, to April 20, 2017); 120 days around the first Trump Administration’s family separation policy (March 15, 2018 to July 15, 2018) 3 ; and the first 90 days of the Biden Administration (January 20, 2021 to April 20, 2021). Contracted care providers make special incident reports to ORR regarding “non-emergency situations that may immediately affect the safety and well-being of a child” (45 CFR § 410.1001; US ORR 2025, Guide to Terms). The anonymized SIRs that ORR produced are not linked with specific anonymized children in the quantitative datasets. Their use here is limited to representative epigraphs of girls in ORR custody disclosing sexual relationships with formal and informal partners.
The quantitative data underlying this study and the SIRs have limitations. The data is administrative and produced by the US government. It is not free from error or bias. Research highlights potential technical and ethical issues with using administrative data generally, including misclassification of important identities and the fact that research participants cannot opt out of data collection (Connelly et al. 2016; Penner and Dodge 2019; Laniyonu and Donahue 2023). With respect to ORR data specifically, the US Senate Finance Committee (2021) issued a scathing report regarding the systemic risks of “abuse and substandard care” for child migrants that ORR’s “extensive record-keeping deficiencies” created.
Neither the quantitative nor the qualitative data in the study present girls’ independent, unmediated voices. The quantitative data includes information the government, not the girl, decided was most relevant. The qualitative SIRs record girls’ experiences also for governmental purposes not the girls’ own and as mediated through the reporting care provider. The limitations of that mediation are clear in a care provider’s express skepticism regarding a girl’s understanding of her relationship with a partner: [Minor] “states she was very happy with her pregnancy and so was her family and her ‘husbands’ family. She explains she had been dating her ‘husband’ as she calls him since they were living together in his mother’s home. . . . he is 22 years old” (SIR 962136). At the same time, the legal processes, including SRIs, that require unaccompanied children to disclose information about sex and relationships may itself be traumatic, even when the adult professionals concerned with the girls’ care do so “to evaluate if they were abused, trafficked, in a forced marriage, or if the pregnancy results from statutory rape” (Ruehs-Navarro 2023; Heidbrink and Díaz 2024, 51). Even with these limitations, the data as a whole, provide important insights into children’s experiences while in ORR custody and ORR’s decision making regarding their release.
This study considered both formal and informal child marriages between an unaccompanied minor and an adult partner. The primary dataset clearly identified a formal marriage where ORR released an unaccompanied child to an adult legal spouse. ORR labeled the relationship between the minor and the sponsor as Category 2 (other family member) with a specific relationship descriptor of “UC’s spouse.”
The sequential order in which ORR produced the data and assigned unique identifiers allowed linking of girl mothers released to their adult husbands with their own biological child(ren) also in ORR custody with them, if any. Generally, a child’s identifier and data rows immediately followed or preceded their mother’s identifier and data rows. CoPilot assisted in developing Excel formulas to link girls released to “UC’s Spouse” with any children also in custody by their sequential placement in the data and the children’s release to “Father.” Manual review verified the assigned mothers and children for congruence across data points including date of entry into ORR custody, program placements, discharge/transfer dates, final discharge type, and, if available, zip code. This study did not account for children to whom ORR assigned a nonsequential unique identifier or placement in the data from their mother.
Informal child marriages were not facially obvious in the primary dataset and required additional coding. This study coded an informal marriage between a girl mother and the adult sponsor to whom she and her child(ren) were released where both of the following were true: (1) ORR categorized the specific relationship between the girl mother and the adult sponsor as “Unrelated Sponsor” and (2) ORR categorized the specific relationship between the girl mother’s child(ren) in ORR custody with her and the adult sponsor to whom they were released as “Father.” As with formally partnered girls, this study linked informally partnered girl mothers with their biological children in ORR custody with them through their sequential presentation in the data and with Excel formulas developed with CoPilot assistance. Manual review verified the assigned mothers and children for congruence regarding dates of entry into ORR custody, program placements, discharge/transfer dates, final discharge type, and, if available, zip code.
Because of data limitations, this study’s coding criteria undercounts informal child marriages for girls in ORR immigration detention. It only accounts for unaccompanied girls in ORR custody with a biological child. It does not account for informal child marriages without biological children, where a girl mother’s biological child had a different father than the adult partner to whom ORR released her, or where ORR assigned “Unrelated Sponsor” to both the girl mother and her child(ren) even where the sponsor was the child(ren)’s father. With limited exception based on application of the verification criteria to the New York Times dataset, it does not account for nonsequential informally partnered mothers and their children also in ORR custody in the primary dataset.
In August 2024, ORR provided explicit policy guidance regarding the assignment of sponsor relationship categories to minor parents and their children also in ORR custody. Where a minor parent has a closer relationship with a sponsor than her child does, ORR now assigns them both the minor parent’s closer relationship at discharge from immigration detention (US ORR 2025, § 2.2.1). ORR’s policy guidance does not address the situation where a child in ORR custody has a closer relationship with the adult sponsor than the child’s minor parent. Assigning a minor parent and her child(ren) also in ORR custody the same relationship with the adult sponsor will continue to obscure informal child marriages in the data.
To calculate the girl mother’s age at the birth of her oldest child with her in ORR custody, this study subtracted the child’s age at entry into ORR custody from the girl mother’s. Because age data is in yearly rather than monthly increments, this calculation provides an approximation of the girl mother’s age at the birth of the oldest child with her in ORR custody.
Results
Demographic Characteristics
During the study period — children admitted to custody between October 1, 2014, and September 30, 2024 and released on or before February 3, 2026—ORR released at least 427 unaccompanied minors from immigration detention to an adult sexual partner. ORR released 201 girls and five boys to their adult spouses. 4 The formally married minors ranged in age from 15 to 17 years. During the same time frame, this study coded 221 informal marriages for girl mothers aged 14 to 17 years and released to their adult partners, 213 in the primary dataset and an additional eight mothers who were nonsequential in the primary dataset but matched the verification criteria in the New York Times dataset.
Countries of Origin
Nearly a quarter of the formally married children (50/206) came from Venezuela, 18.44 percent (38/206) from Afghanistan, 12.62 percent (26/206) from Brazil, and 12.13 percent (25/206) from Cuba. Only 9.71 percent (20/206) came from the Northern Triangle countries of El Salvador, Guatemala, and Honduras. Others came from Bolivia, Burkina Faso, Colombia, Costa Rica, Ecuador, Guinea, Iran, Mauritania, Mexico, Nicaragua, Peru, Russia, Senegal, and Turkey.
Of the minors in informal unions and released to their adult partner, nearly 30 percent (66/221) came from Mexico, 28.51 percent (63/221) from the Northern Triangle countries of El Salvador, Guatemala, and Honduras, and 20.36 percent (45/221) from Venezuela. Others came from Belize, Bolivia, Brazil, Colombia, Cuba, Dominican Republic, Ecuador, Haiti, Italy, Mozambique, Nicaragua, Peru and Romania. None of the informally partnered girls came from Afghanistan.
Average Days in ORR Custody and Placement Types
Overall, the formally married minors averaged 39.70 days in ORR custody with a range from 0 to 239 and a median of 26.50. Formally married minors from Afghanistan averaged 3.63 days in ORR custody with a median of two and a range from 0 to 39 days. Formally married children from countries other than Afghanistan averaged 47.86 days in ORR custody with a median of 32 days and a range from 7 to 239.
Formally married children averaged 1.06 total program placements; 94.17 percent experienced a single placement. ORR first placed 68.93 percent in a shelter. For 17.47 percent — only Afghan minors and all of the Afghan minors — ORR designated the first program placement as Virtual. 5 Only one of the married Afghan minors ever experienced a physical placement. Nearly 10 percent of formally married minors experienced foster care or a therapeutic group home while in ORR custody.
Informally partnered minors averaged 60.34 days in ORR custody with a range from 8 to 380 and a median of 42. They averaged 1.03 total program placements; 96.83 percent experienced a single placement. ORR first placed 75.11 percent in a shelter. 25.33 percent of informally partnered girls experienced foster care or a therapeutic group home while in ORR custody.
Mothers
Fifty formally married girls, nearly a quarter of the formally married minors, were in ORR custody with their own children who ranged from zero to three years old. 6 The girl mothers ranged in age from 13 to 17 at the birth of the oldest child in ORR custody with them. Two had more than one child with them in ORR custody. Four gave birth while in custody as evidenced by a child whose country of birth was the United States of America and date of admittance to ORR custody was after that of the mother.
Formally married girls with children in ORR custody with them were most likely to be from Venezuela (28.00 percent), Guatemala (20.00 percent), and Brazil (18.00 percent). The data shows no formally married girls from Afghanistan in ORR custody with their own children although Afghanistan was the second highest country of origin for formally married minors.
By definition in this study’s coding, all 221 of the informally partnered minors were mothers in ORR custody with their own children. Their infants ranged in age from zero to five years old. The girl mothers ranged in age from 12 to 17 at the birth of the oldest child in ORR custody with them. Eleven had more than one child with them. Five gave birth while in custody.
Release to Spouse or Informal Partner Over Time
Table 1 presents the number of minors ORR released to a formal spouse by the calendar year of the child’s release and the child’s country of birth. The number of minors released to an adult legal spouse ranged from 0 to 51 per year. Figures 1 and 2 represent that same information graphically.
Minors ORR Released to a Formal Spouse by Calendar Year of Release and Country of Birth.
Note: Data for 2014 is for children admitted to ORR custody on or after October 1. Data for 2024 is for children admitted to custody on or before September 30 and released on or before February 3, 2026.
Total Number of Minors ORR Released to a Formal Spouse by Calendar Year of Release.
Minors ORR Released to a Formal Spouse by Calendar Year of Release and Country of Birth.
Table 2 presents the number of minors ORR released to an informal partner by the calendar year of the child’s release and the child’s country of birth. The number of minors released to an adult informal partner ranged from 0 to 112 per year. Figures 3 and 4 represent that same information graphically.
Minors ORR Released to an Informal Partner by Calendar Year of Release and Country of Birth.
Note: Data for 2014 is for children admitted to ORR custody on or after October 1. Data for 2024 is for children admitted to custody on or before September 30 and released on or before February 3, 2026.
Total Number of Minors ORR Released to an Informal Partner by Calendar Year of Release.
Minors ORR Released to an Informal Partner by Calendar Year of Release and Country of Birth.
US States of Release and Age Requirements for Legal Marriage
Zip code data was available for 127 (61.65 percent) of the formally married minors. Based on the sponsor’s zip code, ORR released children to their spouses in 30 states: Arkansas, California, Colorado, Connecticut, Florida, Georgia, Idaho, Illinois, Indiana, Kansas, Kentucky, Louisiana, Maine, Maryland, Massachusetts, Michigan, Missouri, Nebraska, Nevada, New Jersey, New York, Ohio, Oklahoma, Pennsylvania, South Carolina, Tennessee, Texas, Virginia, Washington, and Wisconsin. Florida received nearly a quarter (31/127). See Supplemental Appendix A for age requirements for legal marriage by state and the age range of married minors and years ORR released them.
Zip code data was available for 63 (28.50 percent) of the informally partnered girls. Based on that data, ORR released children to an adult informal partner in 25 states: Alabama, California, Florida, Georgia, Illinois, Indiana, Iowa, Kentucky, Louisiana, Massachusetts, Michigan, Minnesota, Nevada, New Jersey, New York, North Carolina, Ohio, Oklahoma, Oregon, South Carolina, South Dakota, Texas, Virginia, Washington, and Wisconsin. Texas received the most informally partnered minors at 14.28 percent (9/63). See Supplemental Appendix B for statutory rape provisions by state and the age range of informally partnered minors and years ORR released them.
Discussion
During the study period, ORR released at least 427 minors to an adult sexual partner, 206 to a legal spouse and 221 to an informal partner. Because of data constraints that limit the identification of children in informal marriages, this count likely underestimates the number of children the US government released from immigration detention to an adult sexual partner. Whatever the actual count, the decision to release any minor from government custody to an adult sexual partner merits critical attention, attention which this study provides for the first time.
Between 2014 and 2020, the number of unaccompanied children ORR released to a legal spouse fell from a high of 11 to one. In 2021, however, the number jumped to 39, a strong upward trajectory that continued in 2022, 2023, and for children admitted to custody through September 30, 2024. Formal child marriage, which had largely disappeared from ORR data, made a dramatic return.
Legislation prohibiting child marriage in dominant sending countries helps to explain the near disappearance of formal marriages in the data from 2014 to 2020. A unique migration event in Afghanistan explains the dramatic reappearance of formal marriages in 2021. Shifting countries of origin also help explain the continued upward trajectory of children whom ORR released to a formal spouse through the end of the study period in September 2024. Although it is unlikely that the United States will soon experience another unique migration event like the evacuation of Afghans after the Taliban takeover of Kabul in August 2021, shifting countries of origin for migrants arriving to the United States suggest that ORR will also continue to encounter formally married minors in its custody even as the overall number of unaccompanied minors arriving to the United States dropped precipitously in the first year of the second Trump Administration (US ORR 2026).
Between 2014 and 2021, ORR released only six minors to an informal partner, fewer than one per year. That number surged to 31 in 2022 and 112 in 2023, before dropping a third to 72 for the first nine months of 2024. The complicated effects of bans on formal child marriage and the overall prevalence of informal child unions in sending countries suggest that ORR will continue to encounter informally partnered children in its custody. For both formally married and informally partnered children in its immigration detention, ORR faces the challenge of deciding if or when to discharge the minor to an adult sexual partner.
El Salvador, Guatemala, Honduras, and Mexico
El Salvador, Guatemala, Honduras, and Mexico account for well over 90 percent of all unaccompanied migrant children who entered ORR custody during the study period (Alcaraz et al. 2024; US ORR 2026). In 2017, El Salvador, Guatemala, and Honduras all prohibited formal child marriage with no exceptions (Relief Web 2017). El Salvador had previously permitted an individual under 18 to marry a partner with whom they had a child or if pregnant (UNICEF 2017). Guatemala eliminated an exception that had allowed 16- and 17-year-olds to marry with judicial permission (Relief Web 2017). Honduras removed exceptions that allowed minors as young as 16 to marry with parental permission (Moloney 2017). In 2019, Mexico also eliminated exceptions that had allowed children under 18 to marry (Schatzberg 2019).
In determining the legal validity of marriage for immigration purposes generally, US government practice follows the “place of celebration” rule, excepting marriages that violate fundamental public policy such as polygamous unions (USCIS 2025). In other words, the US government will consider a marriage valid for immigration purposes if it was valid where the partners contracted the marriage. US immigration law has only recognized informal marriages in quite limited circumstances involving certain refugees and asylees (American Immigration Lawyers Association, 2025).
ORR has not publicly stated whether it follows the place of celebration rule when assessing the legal validity of marriages for the children in its custody. Assuming, however, that ORR does, the legal bans in El Salvador, Guatemala, Honduras, and Mexico help explain why ORR data record only one girl from these countries released to a legal spouse after mid-2019.
Just as formally married minors from El Salvador, Guatemala, Honduras, and Mexico disappear from ORR data in mid-2019, however, informally partnered girls appear. Of the 129 informally married girls from El Salvador, Guatemala, Honduras, and Mexico in the study, ORR released 128 to their adult partner after mid-2019 when all four countries had banned formal child marriage. Girls from El Salvador, Guatemala, Honduras, and Mexico make up 57.21 percent (123/205) of the surge of minors released to their adult informal partners in 2022, 2023, and 2024. Even before complete bans on formal child marriage in 2017 and 2018, the rate of informal child unions in these countries far outpaced legal marriage. Data from 2015 and 2018 show between 70 percent and 85 percent of ever-married girls aged 15–17 years in informal unions rather than formal marriages in these four countries (UNICEF 2022).
How age limitations on formal marriage affect its prevalence and that of informal unions is complicated (Girls Not Brides 2024). Various researchers have found that changes in the minimum legal age of marriage do not necessarily affect age at first union (Batyra and Pesando 2021; Collin and Talbot 2023; Kidman et al. 2024). In Mexico, the national prohibition on marriage for minors resulted in an increase in informal unions roughly proportionate to the decrease in formal marriage (Bellés-Obrero and Lombardi 2023). No similar studies exist for El Salvador, Guatemala, or Honduras. Data from this study indicate, however, that ORR’s discharge of girls with infants to their adult informal partners increased after El Salvador, Guatemala, Honduras, and Mexico disallowed formal child marriage.
Afghanistan
The Taliban takeover of Kabul in August 2021 sent a wave of Afghans to the United States (United States Department of Homeland Security 2025), including the married girls that entered ORR custody and whom ORR released to their legal spouses. Married girls from Afghanistan make up 92.31 percent (36/39) of the minors ORR released to a spouse in 2021 and thus explain the dramatic reappearance of formal child marriage in the data.
Article 71 of Afghanistan’s Civil Code set the minimum age for girls to marry at 16 but allowed 15-year-olds to marry with their father’s consent or judicial permission (Women and Children Legal Research Foundation [WCLRF] 2008). Despite these age limits, girls younger than 15 are routinely married in Afghanistan and their marriages considered legal (WCLRF 2008). UNICEF data on types of unions show 100 percent of Afghan ever-married girls aged 15–17 years as legally married and none in informal unions (UNICEF 2022). A 2022 study based on a nationally representative survey estimated that “52% of Afghan women aged 20-24 married at ages less than 18 years” (Dardras, Khampaya, and Nakayama 2022).
The form of Afghan migration to the United States at the Taliban takeover of Kabul — emergency evacuation and airlift — also explains unique aspects of married Afghan children’s experiences in ORR custody including their virtual placement and that no married Afghan girls appear in the data with their own children. Augustine-Adams et al. (2025) analyze ORR’s use of virtual program placements that, in the case of Afghan children, allowed ORR to take technical legal, but not physical, custody of the children to enter them into its data management system. Virtual placement, and a special exception to the usual practice in US immigration detention that separates migrant children from accompanying non-parental adults, meant evacuated Afghan children remained physically with non-parental adult caregivers while in ORR’s virtual custody (Augustine-Adams et al. 2025). Thus, married Afghan girls arriving to the United States with a husband but not a parent entered ORR’s virtual custody and likely remained physically with the husband to whom ORR released them.
Virtual custody, where ORR took technical legal but not physical custody of a child, also explains the relative speed with which ORR released formally married Afghan minors to a spouse compared with formally married minors from other countries. Formally married Afghan minors averaged only 3.63 days in ORR custody while formally married minors from all other countries averaged 47.86 days. The median days in custody for formally married Afghan minors was two, with a range from 0 to 39. Formally married minors from all other countries had a median of 32 days and a range from 9 to 239. Virtual custody and the speed with which ORR released formally married Afghan minors to their adult spouses compared to other married girls raises questions about whether ORR sufficiently vetted the adult Afghan spouses.
The data show married Afghan girls arriving to the United States entering ORR’s virtual custody but with no children of their own. It could be that none of the married Afghan minors had children or evacuated with them. The more likely explanation, however, is that any children arriving to the United States with their minor mothers also arrived with their adult fathers. Children arriving to the United States with their adult father would not enter ORR custody, virtually or physically, because they did not meet the legal definition of an unaccompanied minor. Their adult father was available to provide care and physical custody in the United States.
Shifting Countries of Origin
Afghan girls escaping an acute crisis in 2021 does not explain the even higher number of children ORR released to legal spouses in 2022 and the rate at which it continued to do so through the end of the study period. Rather, Venezuelans account for 34.78 percent of the formally married minors released to a spouse from 2022 through the end of the study period, Brazilians for 16.67 percent, Cubans for 14.49 percent, and Nicaraguans for 12.32 percent. The rise in Venezuelan, Cuban, and Nicaraguan minors released to a formal spouse parallels a shift in overall migration from those countries to the United States. Immigration enforcement data show that in FY 2022 and for the first time ever, CBP encountered more Venezuelans, Cubans, and Nicaraguans than individuals from El Salvador, Guatemala, and Honduras, although the primary dataset does not show a similar shift for migrant children (Ruíz Soto 2022; Alcaraz et al. 2024; US ORR 2026).
In contrast to informally partnered girls from El Salvador, Guatemala, Honduras, and Mexico, no distinct pattern emerged in the data with respect to ORR’s release to adult informal partners of girls from Venezuela, Brazil, Cuba, and Nicaragua. During the study period, Brazil, Nicaragua, and Venezuela allowed children under 18 years to marry in certain circumstances (Brazil 2002; Nicaragua 2014; Venezuela 2014). Cuba allowed child marriage until a complete ban in September 2022 (Cuba 1975, 2022). In the 2000s, in Brazil, Cuba, and Nicaragua, the prevalence of informal unions ranged from 50 percent to 70 percent for ever-married girls aged 15–17 years (Barr 2017; UNICEF 2022). Venezuela does not report child marriage data (Barr 2017; UNICEF 2022). During the study period, ORR released Venezuelan minors in relatively similar percentages to a formal spouse and to an informal partner, 52.63 percent compared to 47.37 percent. ORR was more likely to release Nicaraguan and Brazilian minors to a formal spouse at 72.00 percent and 83.80 percent respectively. This study coded only one informal partnership for a Cuban girl mother.
Release of Married Minors to US States with Contrary Age Limitations
During the study period, the legal permissibility of formal child marriage changed in at least two-thirds of the US states to which ORR released married minors to their spouses based on available sponsor zip code data (Supplemental Appendix A). The ORR data do not indicate the date a child married, her age at marriage, or whether she had parental or judicial consent to marry, although this study assumes that, applying the place of celebration rule, ORR considered the marriage to be legal in the country of celebration when it released a minor to a sponsor whom ORR designated as “UC’s Spouse.” Data limitations as well as the changing legal landscape in US states make it difficult to judge, in any particular instance, whether ORR released a married minor to her husband in a state with contrary age limitations.
Where, however, a US state completely banned marriage for minors and allowed no exceptions, ORR’s release of a married girl to her husband there would arguably contravene public policy and, in some circumstances, could make the child marriage void or voidable. Of the documented states to which ORR released married minors during the study period, New Jersey banned child marriage in June 2018, Pennsylvania in May 2020, New York in August 2021, and Massachusetts in July 2022 (Supplemental Appendix A). After the complete bans, ORR released 13 girls to an adult husband in New Jersey, Pennsylvania, and New York. Under specific circumstances, marriages involving an underage partner are void or voidable in each of these three states (Supplemental Appendix C).
Release of Informally Partnered Minors and Statutory Rape Laws
The data do not answer whether, when considering an adult informal partner as the sponsor for a girl mother and her child(ren) in immigration detention, ORR accounted for potential statutory rape in the couple’s prior sexual relationship. As coded in this study, 19 (8.59 percent) of the girl mothers that ORR released to an adult informal partner gave birth between 12 and 14 years of age and another 39 (17.65 percent) at 15 years old.
During the study period and based on the limited available zip code data, ORR appears to have taken some — but not complete — care to release girls to their informal partners in US states where a renewed sexual relationship between the girl mother and her adult informal partner would not constitute statutory rape. ORR released girl mothers who were 16- and 17-years-old on entry into ORR custody to adult informal partners in Alabama, Georgia, Illinois, Indiana, Iowa, Kentucky, Massachusetts, Michigan, Minnesota, Nevada, New Jersey, North Carolina, Ohio, South Carolina, South Dakota, Texas, Washington, and Wisconsin (Supplemental Appendix B). In these states, 16- and 17-year-olds met the minimum age of consent to sex during the study period (Supplemental Appendix B). Fifteen-year olds released in Georgia, however, do not.
ORR’s release of 16- and 17-year-old girls to their adult informal partners in Florida, Louisiana, New York, Oregon, and Virginia is more complicated. These states criminalize sexual conduct with a minor based on age differentials (Supplemental Appendix B). Because the ORR data do not include the age of adult sponsors, this study cannot determine whether the release of 16- and 17-year-old girls to their adult informal partners posed a risk of statutory rape.
California is different. California sets no minimum age for marriage, but 18 as the age of consent to sex (Supplemental Appendices A and B). Absent marriage, California criminalizes all sexual intercourse with a minor; age difference matters only with respect to whether the conduct is a misdemeanor or a felony (Supplemental Appendix B). Thus, when ORR released 16- and 17-year-olds to their adult informal partners in California, the adult partner risked criminal charges of misdemeanor or felony unlawful sexual intercourse (Supplemental Appendix B). Releasing a minor to an adult informal partner runs afoul of California law regarding sex and thus introduces additional precarity to the relationship and to the minor’s well-being.
Conclusion
With an eye to reforming the US detention system for migrant children, this study provides the first critical attention to a US government practice at the juncture of marriage, migration, and childhood: the release from immigration detention of unaccompanied girls to adult sexual partners. The data show that, since 2021, ORR has increasingly recognized child marriages by releasing girls to their husbands and adult informal partners.
In assigning specific final discharge categories to the 427 minors in this study, ORR constructed their formal and informal unions with adult sexual partners as sufficient for the minors’ release to them from immigration detention. The data do not explain, however, what factors made these relationships sufficient — and the release advisable — in contrast to those of other partnered minors whom ORR did not release to an adult sexual partner, especially the thousands of other girl mothers with children also in ORR custody. Nor does the government’s construction of a relationship, or lack thereof, in final discharge categories reveal the minor’s own understanding of it. The epigraphs of this study illustrate possible understandings for partnered girls in immigration detention.
As ORR makes necessarily complex decisions regarding a minor’s release from immigration custody to an adult sexual partner, it should document its consideration of the risks and benefits in a mandatory home study. At present, in-depth home studies are only discretionary “where the sponsor is the partner of the child” (United States Department of Health and Human Services 2024, 45). If ORR does decide to release a minor to an adult sexual partner, it should ensure that the release complies with relevant state law regarding marriage, sex, and consent. Most importantly, as Heidbrink and Díaz (2024) suggest, ORR should develop and “implement robust feedback mechanisms to ensure solicitation, documentation and consideration” of minors’ own views (p. 20). Accounting for minors’ own views regarding their relationships — within the limits of the law — demonstrates respect for their autonomy despite the compounded constraints of childhood, immigration detention, and undocumented legal status.
By documenting the presence of married girls in ORR custody and their release to adult sexual partners, this study begins a conversation. With access to additional qualitative data, further research could provide a greater understanding of ORR’s decision making processes beyond what the primarily quantitative data of this study allow. Further research could likewise assess short, medium, and long-term outcomes for girls whom ORR released to their adult partners, both to support girls’ well-being and to better inform ORR decision making.
Supplemental Material
sj-docx-1-mhs-10.1177_23315024261435093 – Supplemental material for Minor Wives: Married Girls in US Immigration Detention
Supplemental material, sj-docx-1-mhs-10.1177_23315024261435093 for Minor Wives: Married Girls in US Immigration Detention by Kif Augustine-Adams in Journal on Migration and Human Security
Footnotes
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author received no financial support for the research, authorship, and/or publication of this article.
Supplemental Material
Supplemental material for this article is available online.
1
While boys also experience child marriage, the focus here is on girls because of their prevalence in the data underlying this study.
2
In response to FOIA Request 2025-100223-LT-ACF submitted February 15, 2025, ORR produced an Excel spreadsheet with 805,569 rows on January 30, 2026, and an updated spreadsheet with 805,589 rows on February 23, 2026. Each row represents a single placement for a child. Children with multiple placements appear across multiple, sequential rows.
3
4
This study excludes from the total count four infants aged zero to one years whom ORR data identified as released to their spouses.
5
See the Discussion Section for an explanation of ORR’s virtual programs.
6
This study includes in the count of infants, the four aged zero to one year whom ORR identified as released to their spouse.
Federal Statutes and Regulations
United States Code, Title 6, § 279(g)(2)
United States Code, Title 8, § 1232
Code of Federal Regulations, Title 45, § 410
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
